Cell Line Engineering Approaches for the Improved Production of Advanced Biological Medicines

Of utmost importance in the production of any biological medicine is the selection of an optimal producer cell line, such that high-quality, high titer, safe and functional treatments are produced. Volumetric yield remains a challenge in the manufacture of certain recombinant proteins and the drive to engineer mammalian cells towards higher productivity, faster growth and higher product quality is of major interest to the biopharma industry. CRISPR is a revolutionary gene editing tool that can be utilized to knock-out a single gene or, by using a genome-scale guide RNA library, to generate a population of cells with, theoretically, a knockout representing every gene in the genome. A CRISPR/Cas9 genome-wide knockout library was utilized in this work to explore the link between genotype and phenotype in an IgG producing CHO-K1 cell line with the aim of identifying potential gene engineering targets that would be beneficial to cellular productivity. Cells within this library with the highest levels of productivity (top 10%) underwent two rounds of selection by fluorescence-activated cell sorting (FACS) and sequenced to identify guide RNAs that were enriched (525) or depleted (6,290) compared to the unsorted, parental population. From this list, 5 were chosen for validation based on their ranking on the list (extent of depletion or enrichment) and any previous links to productivity-related cellular processes. While no obvious impact on relevant phenotypes was observed in the resulting clones, optimization of critical RNP and FACS selection conditions was achieved resulting in a protocol that can now be used to perform further single knock-out validation studies. The second avenue of host cell line engineering examined was the exploration of HEK293 cell line engineering strategies to optimize rAAV production for gene therapy applications. A combination of directed evolution approaches and physiochemical parameter optimization were explored in a HEK293 suspension cell line producing rAAV. In one particular approach HEK293 cells were transiently transfected with GFP and cells selected based on highest level of GFP expression. These cells were selected, expanded and then used for the production of rAAV in standard triple-transfection process. The resulting virus displayed 30% improved transduction efficiency when applied to target cells. It was also apparent in the various approaches tested that batch to batch variation is a significant challenge in the transient rAAV production system, particularly compared to traditional stable cell line generation for the production of, for example, recombinant proteins. Finally, single cell RNAseq was implemented in a pilot experiment to assess the level of heterogeneity within a population of AAV producing HEK293 cells. The aim was to identify the impact of rAAV production on the HEK293 transcriptome in individual cells within a transiently transfected population, and also to try to measure the level of expression of the various plasmid-derived AAV and helper transcripts in those cells. The most striking learning from this work was that the level of viral genome-encoded transgene, in this case GFP, was so high that the saturation rate achieved in the (short read) library sequencing was only 8%, making it impossible to reliably measure the expression of other transcripts in the samples. Future investigations should either adopt alternative sequencing technologies or establish a means to suppress the levels of GFP present.